cpu.userbenchmark.com
>Sample size of 2 runs
>Both from Australia
>1st run only 1 memory module running at 2666 MHz and 3.85GHz AVG turbo
>2nd run 2 memory modules at 2666 MHz and 4GHz
AMD Ryzen 7 3700X Geekbench result leaked
Cooper Thomas
Other urls found in this thread:
cpu.userbenchmark.com
cpu.userbenchmark.com
userbenchmark.com
userbenchmark.com
userbenchmark.com
userbenchmark.com
userbenchmark.com
cpu.userbenchmark.com
twitter.com
Nathan Rodriguez
why are all the tests on bad memory and lower than max boost
Jeremiah Morris
Elijah Russell
Brayden Howard
There's no way we can know why.
Mason Clark
wow, based AMD
Brody Bennett
>why are all the tests on bad memory and lower than max boost
Max boost is never guaranteed. Try running some prine95 and look at your frequencies yourself.
It _may_ hit max boost on single or double core load if the voltage, temperature and power are in check, but all core boost is often impossible even with overclocking. My 1st gen cant even hit 1core boost frequency on it's own for example.
Cooper Murphy
I hate these benchmarks, they're always ran with suboptimal memory, not even fucking frequency but the timings are always between atrocious or hideous.
That's not even touching the subtimings which are laughable.
You can get way more performance out of a proper memory tune than some 500MHz CPU overclock
Jaxson Williams
>3700X $329
>9700k $370
lol
Cooper Moore
Not interested, these CPU are meant to run with PBO, not manually overclocked.
Kayden Scott
>Max boost is never guaranteed.
you're talking nonsense, look at zen+ results in geekbench, the boosts are much higher on stock
these tests are all run with fucked up bios settings for memory and freq and ONLY THOSE leaked for some reason
utterly bizarre
Connor Wilson
Yeah 3700x is clocked at 4GHz because hurr retardation
Christopher Young
>143 SC score at 4GHz
So it should be 157 at 4.4GHz nice, even better than 9700k oc'd to hell
Asher Mitchell
What do we do now? User benchmark and pissmark gave Intel a advantage but now they've been bought out by AMD, what benchmark does Intel have left?
Angel Lewis
1st gen. Average boost is 3.4. Maximum is 3.6. I may boost on single core test, it won't boost that much on all core. That's where average comes from.
Once again. Open some HWINFO, stress test and check if yourself.
Boost frequency is a marketing meme. If CPUs could hit it all core that easy, everyone would just set it as a base frequency. Beside there's always a lot of binning variations. Not a single CPU is equal when it comes to boost due to variations in voltage requirements and heat.
Owen Mitchell
Pissmark was never good, even if it favors AMD they will "patch" it
Blake Perry
shameful
cope
Henry Ward
Ryzen 3000 switchrs frequency too fast benchmark tools can't figure it out, you're gonna have to use register performance counters to figure out real freq, not Windows slow timers
Eli Evans
t. indiantel
Henry Murphy
Oh yea remember the 3600 pissmark 3 tests?
The only one with 3200mhz and stock advertised actual boost speed was magically deleted shortly afterwards, but the other 2 with fucked up settings stayed up
i said zen+ dipshit
Brandon Reyes
Pure coincidence, go- I mean boy...
Mason Bell
Oliver Harris
also an older single 3600X result - with a core/boost clock of 3.8/3.9 GHz: 143 / 560 / 1,110
Austin Reyes
very nice, maybe good tests will surface later today for the 3700x
Chase Moore
Test with 3733 MHz in 2 days or never?
Chase Hill
Any 3900x's popping up yet?
Brandon Roberts
Wait for 3900x bench at 3.8GHz and 2133mhz ram
Zachary Myers
imagine having two more cores and actually being behind in multithreaded performance
also that price difference
Isaac Brooks
Stop with this antisemitism.
Chase Anderson
Intel has weak pozzed cores.
Ethan Brown
there should be because that is max performance
and even b450 boards can hit that speed
Austin Robinson
>Geekbench
???????
Noah Taylor
>Geekbench
>links userbenchmark.com
Way to go OP
Ryan Cruz
cant wait to get my 3600
so much value !!
Hudson Walker
Pretty nice.
What I'm waiting to see the most with the proper benchmarks is how the boost actually functions. It seems like none of these is boosting near the reported stock clocks and achieving those out of the box shouldn't be impossible by any metric.
That can't be just the benchmarks reporting wrong clocks because this has been the case with all leaks so far.
If we can slap an extra 300-500 MHz to all of these results, we're looking at pretty incredible SC scores. Can't wait to see the real benches.
Maybe it's some loophole they're trying to exploit to avoid getting raped in court if AMD sues them for breaking the NDA.
>The NDA said we aren't allowed to test the products to their fullest potential, see we're not breaking the NDA! The CPU is gimped, don't sue us pls.
It's either that or somehow everyone has fucked up BIOS that doesn't allow for boost clocks and decent RAM settings.
Nathaniel Gutierrez
If anything, it's stores getting stock in on a friday and staff buying and building their own new system before the official release dates.
Justin Young
They can also come from mobo vendors as well testing non standard settings en masse where various test setups accidentally upload their results.
David Gutierrez
you people are pathetic and sad and pathetically sad
Zen3000 clocks low because there's twice as many transistors in the same space. 2 times the xtors @ 50% xtor power = 1.5x thermal density of a zen2000 cpu
ya'll motherfuckers believed the hype.
it's like you're all 14 and mentally disabled
Nathan Clark
What the hell is Zen 3000 and 2000? Did you hit your head? Are you okay, user?
Julian Diaz
Some real numbers
Joshua Gray
He was merely pretending xdDD
Adrian Garcia
5.3 vs 4.0
Intel is done
Dylan Torres
checked
tech "experts" btfo
Brayden White
this. it is literally over.
Robert Clark
nobody uses intel past 5th gen in any serious capacity